The portal internal links seem to work up to the point of clicking the Dashboard settings, then the links break. The links the portal generates after running settings no longer have the api-id in them so they fail.
Is there a fix for this or am I doing something wrong?
We just ran through the whole process here and can confirm that the docker container works as expected. Though, this being docker, there are always quirks.
How are you running docker? Does it spin up on 127.0.0.1 or does it provide a bridged host IP? You will need to make sure that:
You have set the correct internal IP in the three greyed out proxied APIs to that of the portal (they default to localhost, but this can fail depending on the install)
You have set the hostname in the “Your developer portal” -> “Set your portal domain” to be www.tyk-portal-test.com
You ahve only run the setup.sh once (it can be run more than once but this can cause issues)
I’ll work with our ops guys tomorrow to see what they find. They run Docker a lot.
I deleted and recreated dashboard and mongo containers and ran setup again (deleting the mongo container I think make it safe). The set portal domain now completes successfully, but now I get
{
“error”: “There was a problem proxying the request”
}
Our Ops guys are getting exactly the same behaviour. We are either following the same incorrect procedure, both have something else on our machines that is interfering, or it is broken.
We’ve put together another docker container and validates it against docker on OSX as well as Docker on windows, to use t, clone the repository and then follow the steps in the readme file:
Great! Thank you! That works, except that if I try to log on with the existing [email protected] user I get a CRSF error, but registering a new user and logging in with that credential works.